Structure and Properties of ZnO-ZnS HeterostructuresStructure and Properties of ZnO-ZnS Heterostructures Structure and Properties of ZnO-ZnS Heterostructures
Sanchali Das Missouri State University, Sanchali026@live.missouristate.edu
As with any intellectual project, the content and views expressed in this thesis may be
considered objectionable by some readers. However, this student-scholar’s work has been
judged to have academic value by the student’s thesis committee members trained in the
discipline. The content and views expressed in this thesis are those of the student-scholar and
are not endorsed by Missouri State University, its Graduate College, or its employees.
Follow this and additional works at: https://bearworks.missouristate.edu/theses
Part of the Semiconductor and Optical Materials Commons
Recommended Citation Recommended Citation Das, Sanchali, "Structure and Properties of ZnO-ZnS Heterostructures" (2019). MSU Graduate Theses. 3422. https://bearworks.missouristate.edu/theses/3422
This article or document was made available through BearWorks, the institutional repository of Missouri State University. The work contained in it may be protected by copyright and require permission of the copyright holder for reuse or redistribution. For more information, please contact BearWorks@library.missouristate.edu.
A Master’s Thesis
Master of Science, Material Science
By
iii
Physics, Astronomy and Material Science
Missouri State University, August 2019
Master of Science
ABSTRACT
Heterostructures consisting of at least two layers of dissimilar materials have always been well
studied due to potential applications in nano-electronic and optoelectronic devices. In this study,
I have investigated the structural and physical properties of zinc oxide -zinc sulfide (ZnO-ZnS)
based heterostructures of thin films and core-shell nanoparticles. Pulsed laser deposition
technique was employed to grow ZnO thin film on sapphire substrate. ZnO-ZnS heterostructures
were obtained using hydrothermal synthesis where thiourea has served as a precursor solution for
the source of sulfur. X-ray diffraction analysis on the parent and the sulfidized samples gives the
information about the crystallinity and the phase of the heterostructures. Scanning electron
microscope analysis provides the morphological and constituent elemental information of the
heterostructures. Variation of the time and temperature of the sulfidation can alter the structure
and properties of the heterostructures. At a given temperature, the amount of sulfidation depends
on the exposure time on the sulfur rich environment. The more the availability of sulfur ions due
to thiourea decomposition, the more is the growth of ZnS at the ZnO-ZnS interface. Raman
spectroscopy gives the vibrational states present in the both parent and synthesized
heterostructures while photoluminescence analysis provides information about optical band gap
and defect states that are present in the heterostructures.
KEYWORDS: heterostructures, optoelectronics, crystallinity, vibrational states, band gap,
photoluminescence, hydrothermal synthesis
By
Of Missouri State University
For the Degree of Master of Science, Material Science
August 2019
David Cornelison, Ph.D., Committee Member
Ridwan Sakidja, Ph.D., Committee Member
Tiglet Besara, Ph.D., Committee Member
Julie Masterson, Ph.D., Dean of the Graduate College
In the interest of academic freedom and the principle of free speech, approval of this thesis
indicates the format is acceptable and meets the academic criteria for the discipline as
determined by the faculty that constitute the thesis committee. The content and views expressed
in this thesis are those of the student-scholar and are not endorsed by Missouri State University,
its Graduate College, or its employees.
v
ACKNOWLEDGEMENTS
I would like to thank my academic and research advisor Dr. Kartik C. Ghosh for immense
support and guidance through two years of my research. He has been encouraging to take some
novel approach while conducting the research. I am also grateful to professor Dr. David
Cornelison for advising and helping me in critical thinking while doing the experiments. I thank
my other committee members Dr. Ridwan Sakidja and Dr. Tiglet Besara for their insightful
opinion.
I appreciate the help of my other group partner Christopher Robledo for characterization of my
samples and continuous support. I thank my other friends Nazmul Alam and Nirmal Baishnab
for their support.
I dedicate this thesis to my parents and my husband.
vi
Structural and Optical Properties of Heterostructures Of ZnO-ZnS Thin Films
Page 10
Conclusions Page 47
Heterostructures
Conclusions Page 65
LIST OF TABLES
Table 1: FWHM and crystallite size calculation for porous ZnO parent films Page 16
Table 2: Lattice parameter of the porous ZnO parent films Page 18
Table 3: Lattice parameters of the sulfidized films Page 24
Table 4: Crystallite size of the sulfidized samples from porous ZnO Page 24
Table 5: Lattice parameter of non-porous parent ZnO sample Page 28
Table 6: Lattice parameter of ZnS sample 12 sulfidized for different amount of time Page 28
Table 7: Crystallite size of non-porous sulfidized ZnO sample Page 28
Table 8: Experimental and reported value of Raman active modes for porous ZnO
thin film
Page 36
Table 9: Comparison of the Raman active modes of reported and experimental data
of ZnS
Page 37
viii
Figure 4: ZnO-ZnS thin film heterostructure Page 8
Figure 5: ZnO-ZnS core-shell heterostructure Page 8
Figure 6: Migration of atoms via vacancies in lattice Page 9
Figure 7: Flow chart diagram of the synthesis of ZnO-ZnS thin film
heterostructures
Figure 8: XRD data of porous ZnO parent Page 16
Figure 9: Cif data of wurtzite ZnO Page 17
Figure 10: a) XRD of sulfidized sample (16hr) b) cif file of ZnS Page 19
Figure 11: XRD of sulfidized sample (4 hour) (a) sample1 (b) sample2 Page 20
Figure 12: XRD of sulfidized sample (1hr) Page 21
Figure 13: Overlapping plot of parent and sulfidized samples for different time
duration
Page 21
Figure 14: Lattice parameter vs sulfur mole percent (a) 16 hours (b) 1 hour
sulfidized films
Page 23
Figure 15: a) XRD of nonporous parent ZnO, b) 16 hours sulfidized ZnO, c) 32
hours sulfidized ZnO, and d) 1 hour sulfidized ZnO
Page 26
Figure 16: Comparison of sulfidation for porous and non-porous film for 16 hours Page 27
ix
Figure 17: Overlapping of XRD plots of sulfidation of non-porous ZnO films for
different time duration
Page 27
Figure 18: SEM micrographs of (a) parent sample and 16 hour sulfidized sample
(b) (40µm) (c) 10µm (d) 20µm
Page 29
Figure 19: EDS spectra of the sulfidized sample Page 30
Figure 20: Raman active modes of ZnO nanoparticles Page 31
Figure 21: Overlapping plot of porous ZnO film with the ZnO Nano powder Page 32
Figure 22: Deconvolution of peak of 574 cm1 Page 33
Figure 23: Raman active modes of non-porous ZnO Page 33
Figure 24: Raman active mode of sulfidized sample for 16 hours Page 34
Figure 25: Deconvolution of Raman active modes (LO/SO) Page 35
Figure 26: Raman spectra of ZnS sample (sulfidized for 1 hour) Page 35
Figure 27: PL spectra of porous ZnO thin film Page 38
Figure 28: PL spectra of non-porous ZnO thin film Page 38
Figure 29: PL spectra of porous ZnO sample 3 Page 40
Figure 30: PL spectra of bulk ZnO sample Page 40
Figure 31: PL spectra of the film sulfidized for 16 hours Page 41
Figure 32: PL spectra of the porous sulfidized film (1 hour) Page 42
Figure 33: PL spectra of ZnS sample12(16 hour sulfidized) Page 43
Figure 34: PL spectra of ZnS sample12(32 hour sulfidized) Page 43
Figure 35: PL spectra of sample12 (1 hour sulfidized) (a) UV emission and (b)
defect emission
Page 44
x
Figure 36: (a) XRD of a porous parent ZnO (b) overlapping plots of 4 hours
sulfidation at 180°C
Page 46
Figure 37: Mechanism of ZnS growth during ZnO sulfidation of inward and
outward growth
Page 51
Figure 38: XRD of wurtzite ZnO nano powder Page 53
Figure 39: Cif data of wurtzite ZnS Page 54
Figure 40: XRD of ZnS nano powder sulfidized for 16 hours Page 55
Figure 41: XRD of ZnS nano powder (6 hours sulfidized) Page 56
Figure 42: XRD of ZnS nano powder (1 hour sulfidized) Page 56
Figure 43: a) Overlapping XRD plot of ZnO nano powders sulfidized for different
hours b) comparison with ZnS cif data
Page 57
Figure 44: SEM micrograph of (a) pure ZnO nanoparticle, (b) ZnS nanoparticle
(1 hour sulfidized), (c) 6 hours, and (d) 16 hours
Page 58
Figure 45: EDS analysis of sulfidized NPs for (a)16 hours and (b) 6 hours Page 59
Figure 46: Raman active modes of sulfidized NPs for (a)16 hours, (b) 6 hours, and
(c) 1 hour
Page 60
Figure 47: PL spectra of 16 hours sulfidized nanoparticles Page 62
Figure 48: PL spectra of ZnO NP Page 62
Figure 49: Defect states of ZnS ZB Page 63
Figure 50: PL of ZnS nanoparticles (6 hours sulfidized) Page 64
Figure 51: PL of ZnS nanoparticles (1 hour sulfidized) Page 64
1
INTRODUCTION
A heterostructure is a sandwich structure consisting of two dissimilar materials.
Generally, semiconductor heterostructures made of two dissimilar semiconductors having
different band gap values can create unique structural and electrical properties.1 The difference in
the band gap enables spatial confinement of injected electrons and holes. The difference in the
refractive index between two dissimilar semiconductors can aid forming waveguide.2
Heterostructures have interesting properties like confined carrier motion that enhances carrier
mobility in doped semiconductor. Heterostructures have applications in several devices like laser
diode, light emitting diode, solar cell, optical detectors, high electron mobility transistors
(HEMTS).3
Most popular semiconductors exploit III-V and IV-VI compounds with similar crystal
structures and less mismatched lattice constant.1 Researchers have studied GaAs and AlxGa1-xAs
extensively along with other heterostructures.4 A close lattice matching is necessary for growth
of epitaxial layers of heterostructures for prevention of the carrier recombination at the
heterojunction interface. As there is growing interest in the field of thin film and
nanotechnology, ultrathin multilayer heterostructures are being studied more extensively.
Quantum confinement is one of the intriguing phenomena in the heterostructure when narrow
energy band gap layer becomes atomically thin such as few tens of nanometers or even less than
that.5 Especially, this becomes more pronounced when the layer or the film dimension becomes
comparable with De-Broglie wavelength. This ultrathin heterostructures are known as
superlattices. As the size of materials approaches nanoscale range (less than 100 nm) surface
properties become more prominent compared to the bulk properties. Same goes true for ultra-thin
2
or nanostructure heterojunction. High surface to volume ratio of the nanoscale materials gives
rise to unique novel properties.
Prevalent methods of epitaxial growth mechanism such as molecular beam epitaxy
(MBE) and Metal organic chemical vapor deposition (MOCVD) have enabled fabrication of
wide variety of thin layers of semiconductor compounds on various single crystalline substrates.6
These heterostructures based devices can be classified into two groups: Heterojunction and
Quantum Well.
Heterojunctions are comprised of two layers A and B, and electrons or holes exist
between the interface layers. The quantum well can be represented in the form of ABA where the
central layer may confine charge carriers if the system is doped. For both heterojunction and
quantum well, charge carriers are trapped in two dimensions. High electron mobility and strong
electron-electron interaction are the result of electron confinement. The restriction of the charge
carriers (electron, hole or exciton) in the direction normal to the film or layer can be considered
carrier confinement in one dimension. For thin films, well can be considered as deep infinite
square well. For semiconductor heterojunction the potential wells are finite and rectangular in
profile.
Superlattices are comprised of multiple periodic array of semiconducting compounds in
the form of ABAB layers.6 Multiple quantum well of composition ABAB…contains wide
barriers such that carriers are confined in separate well. While the characteristics of a superlattice
can be approximated by Kronig-Penny model where role of the interface is very important for
the electronic properties.
Most of the active photonic devices utilize III-V semiconductor heterostructures where
the optical properties are very important. Some of the properties like spontaneous emission,
3
absorption spectra, and refractive index, despite having been studied for decades by researchers,
it has remained hard to approach device modelling.7 It is necessary to study the microscopic
details of light matter interaction in III-V semiconductor for both bulk and nanomaterial for a
multilayer device, especially how carrier injection is utilized for doped or undoped layers.
Previous researchers reported that for direct band gap semiconductors like GaAs or InP Kane
model is being used for approximation. It considers conduction band ( valley) with the heavy-
hole, light hole and split-off valence band. X and L valleys play an important role to determine
electron Fermi level and intra-band absorption co-efficient. After the electron states and carrier
distribution of those are determined, the rate of radiative transition is calculated using
perturbation theory.7 For the calculation of the first order process six types of direct radiative
recombination are considered. For second order process, scattering mechanisms were
incorporated approximating three parabolic sub-band valence band and non-parabolic
multivalley conduction band.
ZnO and ZnS are popular wide band-gap group II-VI semiconductors used for fabricating
optoelectronic and photonic devices. Heterostructures of wurtzite ZnO (band gap 3.37ev) and
wurtzite ZnS (EG ~3.72ev) or zinc-blende ZnS (EG ~3.7ev)8 have applications in ultraviolet laser
and photovoltaic detectors.
Yan et al.8 reported synthesis of biaxial ZnO/ZnS nano heterostructure which has been
challenging till this date. They investigated growth and structure of 1D heterostructure
comprising of II-VI semiconductors. They also studied excitonic properties of these structures by
employing cathodoluminescence spectroscopy. They synthesized hetero-crystalline ZnS /single
crystalline ZnO based biaxial nanobelts by thermal evaporation method where gold is used as
catalyst. The TEM image revealed uniform interface between the upper ZnS and lower ZnO
4
fragment. The individual diameters were reported to be 40nm and 30nm in individual ZnS and
ZnO domain. They reported that the ZnO fragment was wurtzite in structure and ZnS portion
comprised of hetero-crystalline superlattice structure of alternative zinc blende (ZB) and wurtzite
(WZ) ZnS sections. For investigation of the optical properties of the heterostructure,
cathodoluminescence (CL) was studied on the sample. A broad emission spectrum was reported
around ~ 535 nm and a weak UV emission band was centered around~ 345 or ~385 nm. This
result was consistent with the reported morphology of ZnS/ZnO nano heterostructure.
Heterostructure device consisting of two dissimilar materials with unique structural and
optoelectronic properties can exhibit enhanced performance such as emission efficiency or
distinct luminescence properties. ZnS-In based heterostructure was reported to show enhanced
emission efficiency, SnO2-ZnO heterostructure exhibited unique luminescence properties.9 Three
dimensional ZnO/Si based heterojunction improves the efficiency for solar water splitting. Tian
et al.9 previously reported ZnS nanobelt based UV-light sensors which, however, shows weak
photocurrent and poor stability. ZnO nanostructure based UV photodetectors exhibits poor
photocurrent stability and slow responsivity mostly due to internal defects like oxygen vacancies
or zinc interstitials.9 However, it has been established that highly crystalline ZnS based
photodetectors have faster response and speed while ZnO with good contacts can generate large
photocurrent. By combining these unique properties, they fabricated hybrid ZnO-ZnS based
photodetector which is potentially promising for UV-light sensors. This fabricated device has
shown fast response speed, tunable spectral selectivity and enhanced photoresponsivity. The
ZnO-ZnS heterostructure have shown biphase mixture of wurtzite ZnO and wurtzite ZnS with
low level impurity. Two luminescence spectra have been observed at ~ 396nm and ~530 nm for
ZnO-ZnS hybrid structure. The green emission spectra around 530 nm resulted from ZnO
5
nanorod within the hybrid structure. According to this paper, the linear I-V characteristic curve
was reported which indicates good Ohmic contacts between the electrodes and the film. The
device was illuminated with light sources of different wavelength of 300, 320,350,380 nm. The
increase in photocurrent was reported for the excitation wavelength below 380nm. This is due to
increased number of electron-hole pairs when excitation energy is greater than the band gap of
the ZnO-ZnS heterostructure.
Lu et al.3 reported vertically aligned ZnO-ZnS heterojunction nanowires, synthesized by
thermal evaporation in presence of residual oxygen. These were used to convert mechanical
energy to electrical energy. Vertical ZnS nanowire arrays were synthesized by selective etching
ZnO-ZnS nanowire arrays. Wurtzite structure of both ZnO and ZnS were being reported. TEM
images have shown epitaxial growth of ZnO above ZnS nanowires at the interface. Large lattice
mismatch between ZnO and ZnS can result into thin intermediate thin layer in between. Both
TEM and EDS results have confirmed ZnO-ZnS heterojunction nanowire array. The
piezoelectric property was examined by using atomic force microscopy (AFM ).3 CL
measurement of ZnO-ZnS array has shown emission around 509 nm. For this paper, the goal was
to synthesize ZnO-ZnS heterostructure and study the structural and optical properties of the
nano-heterostructure. ZnO being a wide direct band gap semiconductor with high exciton
binding energy (60 mev) has good thermal and chemical stability.10 ZnO has been well studied
for divergent applications in different fields such as piezoelectric material, UV light emitting
diode, gas sensing, transparent conductive oxide (TCO) layer in thin film based solar cell, and
transducer.10 Several studies have proved the presence of defect sites such as oxygen vacancies,
oxygen interstitial, and zinc interstitial and so on which gives rise to the defect related band in
photoluminescence spectrum. Zinc oxide crystallite has two different crystal structure i) wurtzite
6
and ii) zinc blende.11 Wurtzite is the most stable structure where each anion is bonded with four
cations at the corner of a tetrahedron.11 Figure 1 represents the structure of wurtzite.
Zinc sulfide (ZnS) is a popular wide band gap (3.7ev) II-VI semiconductor material for wide
variety of application such as light emitting diode, optical coating, electroluminescent devices,
heterojunction solar cell12 and so on. Effective phosphor material also fabricated using ZnS thin
films doped with transition metal.13 Zinc sulfide exists in two different polymorphs. i)
Zincblende (sphalerite) and ii) wurtzite. In zinc blende structure, the sulfur atoms occupy the fcc
sites where zinc atoms occupy half of the tetrahedral sites. This structure is also considered as
two inter-twined fcc lattice corresponding to each sulfur and zinc with the origin displaced at
half of the body diagonal. The reported lattice parameter is a= 0.541 nm. In wurtzite structure the
sulfur ions arrange themselves in HCP array and half of the tetrahedral sites are occupied by zinc
Fig 1: Structure of wurtzite ZnO Figure 1: Unit cell of wurtzite ZnO
7
ions. In this study, thin film heterostructure of ZnO/ZnS was fabricated using pulse laser
deposition (PLD) technique and hydrothermal method. Also, we synthesized sulfur doped nano
heterostructure of ZnO/ZnS core-shell nanoparticles. Figures 2 and 3 represent the unit cell
structure of zinc-blende and wurtzite ZnS.
The structural properties were studied using XRD, SEM and EDS. Understanding the structural
and physical properties of ZnO-ZnS based heterostructures can lead to develop efficient
optoelectronic devices. Figures 4 and 5 represent the schematics of ZnO-ZnS thin film and
CSNP based heterostructure.
Very few studies have been done on the sulfidation of ZnO thin film and development of
the ZnO/ZnS heterostructures. Zhang et al.14 reported sulfidation of ZnO thin film by annealing
at 500°C in H2S gas mixture. According to their study, H2S gas is highly reactive and easily
controllable for sulfidation of ZnO thin film compared to sulfur vapor. However, the underlying
Figure 3: Unit cell of wurtzite ZnS Figure 2: Unit cell of zinc blende ZnS
8
mechanism of sulfidation or the physics behind it, is not very clear from their study. Bezverkhyy
et al.15conducted sulfidation of pure and Cu-doped ZnO nanoparticle at 250 and 350°C using
gaseous H2S. The resulted structure was studied through transmission electron microscopy
(TEM) and in situ synchrotron XRD. Partial sulfidation of ZnO with H2S has resulted formation
of cavities at the ZnO/ZnS interface which confirmed sulfidation of ZnO accompanied through
outward growth of ZnS.
Figure 5: ZnO-ZnS core-shell heterostructure
9
This phenomenon is known as Kirkendall effect which proved Zn2+ and O2- ions migrated
to the external ZnS/ H2S interface. For Cu-doped ZnO inward growth process was observed
which depicts diffusion of S2- anions from external H2S gaseous phase to internal ZnO/ZnS
interface where exchange of anions take place. In our experiment, exchange of anions in ZnO
happens from thermal decomposition of aqueous solution of thiourea which serves as a source of
sulfur. In addition, with XRD, Raman, PL and SEM analysis, TEM micrograph needs to be
analyzed to observe formation of voids at the interface of ZnO/ZnS nanostructure. The diffusion
mechanism is strongly dependent on activation temperature and time. Figure 6 represents
migration of atoms through lattice via vacancy formation.
Proper temperature can accelerate the migration of anions and cations through crystal
structure. Hence vacancy related defect in parent crystal structure will be helpful for faster
diffusion of S2- anions to be exchanged with O2-.
Figure 6: Migration of atoms via vacancies in lattice
10
ZnS THIN FILMS
C-axis oriented ZnO-ZnS thin film heterostructures have been synthesized by
sulfidation of the pulsed laser deposited ZnO thin films using thiourea solution in
hydrothermal synthesis. The PLD grown ZnO films were sulfidized at 180°C for varying
amounts of time. The parent ZnO and the sulfidized films were characterized by x-ray
diffraction, scanning electron microscopy (SEM) and energy dispersive electron
spectroscopy (EDS). Optical properties were studied through Raman and
photoluminescence spectroscopy. It turns out through structural analysis that crystallinity
and preferred C-axis orientation decreases with initial sulfidation and regain with
sulfidation time. We also found that porosity of the film enhances the exchange of
oxygen to sulfur ion. PL spectroscopy data provide information about the defect related
states present in the band gap of pure ZnO and the sulfidized films. More exposure time
in the sulfur environment showed better and prominent sulfidation retaining the wurtzite
crystal structure. Raman spectroscopy provides information about the vibrational modes
present in both parent and the sulfidized films.
Introduction
ZnS being a high direct band gap semiconductor (3.54ev-cubic/ 3.91ev-WZ)16 has
potential applications in numerous optoelectronic devices such as light emitting diode, solar
cell14, electronic displays, and sensors.17 Several synthesis methods are popular to fabricate ZnS
11
thin films such as chemical bath deposition (CBD),sol-gel, chemical vapor deposition (CVD),
magnetron sputtering, and molecular beam epitaxy.14 Very few papers reported synthesis of ZnS
thin films by PLD method using a ZnS target. In most of the reported papers, ZnS thin film has
been deposited on preferred substrate by magnetron sputtering method using a compound
sputtering target.14 Some other papers reported CBD method as large area and cost-effective
production of ZnS thin film in relatively low temperature. But in general, the CBD grown films
are either amorphous or not very crystalline in nature.18 Therefore , we have approached a
different path to synthesis ZnS thin film using the combination of pulse laser deposition(PLD)
and hydrothermal synthesis. In this paper, ZnO-ZnS thin films are fabricated by sulfidation of
ZnO thin films prepared by PLD. Even though sulfidation of ZnO thin film has been reported
previously by Zhang et al.14, their methodology was different. According to that paper, ZnO thin
films were deposited by reactive sputtering from a Zinc (Zn) target. These ZnO films were
sulfidized at 500°C in sulfur vapor. Influence of RF power and effect of sulfidation temperature
were studied to understand the properties of ZnS thin films. Another research group reported
annealing of ZnO thin films prepared by spray pyrolysis method in presence of sulfur powder at
450°C.19
In my experiment, I deposited ZnO thin film on sapphire substrate by PLD method and
the sulfidation reaction was carried out by hydrothermal synthesis using a precursor solution of
thiourea which acts as a source of sulfur. Good quality ZnO films were developed via PLD
which resulted in moderate deposition rate of the species and appropriate temperature was
helpful for thermal mobility and adhesion of species on the substrate. The oxygen in the metal
oxide films were replaced by sulfur by diffusion mechanism of the sulfur atoms. After the
growth, the structural and optical properties of the resultant ZnS films were analyzed.
12
Experimental Methods
Figure 7 represents flow chart diagram of synthesis process of ZnO -ZnS thin film
heterostructure.
Figure 7: Flow chart diagram of the synthesis of ZnO-ZnS thin film heterostructures.
C-axis oriented porous ZnO films were grown on single crystal (0001) sapphire substrates (α-
Al2O3) by PLD. Both ceramic ZnO and metallic zinc target were used for laser ablation. The
ceramic ZnO target (diameter 1.00"×0.25"thick) was prepared by conventional pressed method
at 100 Mpa pressure using 99.99% pure ZnO nano powder which was sintered at 1300°C in air.
To carry out the deposition first the stainless-steel vacuum chamber was evacuated to the base
pressure of 6.5×10-5 mbar using a turbo molecular pump. A solid-state pulsed Nd: YAG laser
(wavelength 266nm) was focused on the rotating Zn or ZnO target to ablate the species. During
13
deposition the chamber was filled up with high purity oxygen gas with varying pressure of 4×10-
2 to 4 x10-1 mbar. The target to substrate distance was kept at 35mm and deposition temperature
was maintained at 600°C to enhance the mobility of the deposited species on the substrate. The
laser energy density was 2.54J/cm2. 500 laser shots were used to pre-clean the target and 20,000
shots were given for laser ablation for the duration of 45 minutes (repetition rate 10Hz, Q-switch
delay-5µs). After deposition, ZnO films were annealed in oxygen atmosphere of 1.1×10-1 mbar
pressure at 600°C for an hour. The average film thickness measured after annealing by Veeco,
Dektak was around 140nm.
The as-deposited porous ZnO films were sulfidized by hydrothermal synthesis where
thiourea dissolved in DI water served as precursor solution. Different As grown ZnO thin films
were immersed in .1M thiourea aqueous solution (prepared by dissolving 133mg thiourea in 17.5
ml DI water) inside an autoclave which was heated at 180°C for 16-hour, 4-hour and 1-hour
respectively. After sulfidation, the autoclave was cooled down to room temperature and the
sulfidized films were thoroughly rinsed with de-ionized water.
The sulfidation mechanism inside the autoclave took place after the thermal
decomposition of thiourea [CS(NH2)2] in aqueous solution. A possible reaction could be as
below.
OH- + CS(NH2)2 → SH- + CO(NH2)2 (urea)
SH- ion could serve as source of S2- ions which can exchange with O2- in ZnO.
In general, the reaction can be represented as the following steps below.
H2O + CS(NH2)2 →H2S + CO(NH2)2 (urea)
14
ZnO + H2S → ZnS + H2O
The crystallinity and the phase of the as-deposited ZnO films and the sulfidized samples
were characterized by x-ray diffractometer (Bruker AXS D8) equipped with high-resolution
detector using Cu-Kα source with wavelength of 1. 5406.The operating voltage and current
were 40 kv and 40 mA, respectively. The θ-2θ scan was performed from 20° to 80° after
optimization of and Z-axis scan. The data were analyzed using Origin pro 8.5.
The surface morphology of the as-deposited porous ZnO films and the ZnO-ZnS
heterostructure films were studied through a scanning electron microscope (FEI Quanta 200). An
energy dispersive x-ray spectroscopy detector (EDS) equipped within the same system was used
for elemental analysis of the samples. SEM gives the information about the size and shape of the
material. The microscope was operated at 10KV accelerating voltage while the working distance
was fixed at 10 mm, measured spot size was 3.0. The EDS acquisition time was 120sec and
performed by selecting a rectangular area on the sample. Several spots were studied to get better
statistical measurement of the sample. Conductive gold or carbon coating was sputtered on the
samples before performing the SEM-EDS measurement because the sapphire substrate is an
insulating material which can result into charge build up on the sample surface that can create
band bending resulting in distortion in image.
The Raman spectroscopy technique was employed to probe the information about
vibrational energy states present in the crystal. The basic principle of Raman spectroscopy is
based on inelastic scattering of incident excitation through the interaction of atomic and
molecular vibration of the sample. As the intensity of anti-stokes scattering is 100 times less than
stokes-scattering, stokes scattering was studied. All ZnO-ZnS heterostructures were
characterized using 532nm green laser by Horiba Labram HR Raman PL system.10x objective
15
lens was used for initial focusing. 50x lens was helpful to determine more localized spot. For
photoluminescence spectroscopy, the samples were irradiated with 325 nm He-Cd laser source at
room temperature. Excitation energy was above the band gap of the material. The samples were
placed on a metal bar of tungsten carbide which contributes minimum background signal. The
15x NUV-lens was used for focusing on the localized spot and measurements were taken at
multiple spots. Beam splitter UV splits the scattered light from the sample according to the
wavelength. The scan was taken in the range of 350-750 nm while acquisition time was 20
seconds, and 20 scans were taken in average in that wavelength range. Both Raman and PL
spectroscopy data were analyzed using origin pro 8 software.
Results and Discussion
Figure 8 represents XRD patterns of as grown and annealed porous ZnO thin films.
Several samples were synthesized via pulsed laser deposition. A strong diffraction is observed at
2θ= ~34.5° which corresponds to hexagonal wurtzite phase (002) and a weak diffraction peak
appears at 2θ= ~72.5° which corresponds to (004) phase. It is evident that all ZnO films have
preferred orientation along the C-axis perpendicular to the sapphire (0001) substrate because of
lowest surface free energy along that direction.20 Diffraction peaks of two parent ZnO films are
represented in Figure 8. The crystalline quality can be interpreted from the full width half-
maximum and intensity of ZnO (002) peak. Table 1 represents the values of FWHM for two
different parent films . Smaller value of FWHM indicates better crystalline qualities of the
synthesized film. It also indicates higher crystallite size of the PLD grown films. As calculated in
Table 1, the narrow FWHM and higher intensity of XRD peaks depict better crystalline quality
of the films annealed in O2 atmosphere. The crystallite size can be calculated using Debye-
16
Scherrer formula as given below.
Table 1: FWHM and crystallite size calculation for porous ZnO parent films
From the above Table, I observed higher crystallite size of sample 1 compared to sample 4
which depicts higher crystallite growth of parent sample 1 after annealing in oxygen atmosphere.
ZnO
Parent
Sample
K FWHM λ(Å) (rad) cos(rad) ß(rad) L(nm) h k L
1
0.9 0.64917 1.5406 0.3016 0.954 0.011330154 12.82 0 0 2
4
0.9 0.8302 1.5406 0.298 0.956 0.014489723 8.7 0 0 2
Figure 8: XRD data of porous ZnO parent
17
This might have resulted from variation of the background oxygen pressure during synthesis and
growth of the parent ZnO films. The .Cif file gives confirmation of the crystallographic
orientation in figure 9.
= dimensionless shape factor 0.9
= line broadening at half of the maximum intensity (FWHM) in radian
= Bragg angle
Figure 9: Cif data of wurtzite ZnO
18
Table 2: Lattice parameter of the porous ZnO parent films
From Table 2 ,it can be seen that the lattice parameters got matched well with the theoretical
standard values of hexagonal ZnO (a= 3.1249 and c= 5.20 6).21 Sulfidation was performed on
different parent ZnO film for variable amount of time i.e 16 hours ,4 hours and 1 hour. Figure
10: a) and b) shows the XRD of the sample which was sulfidized for 16 hours and the .cif file for
wurtzite ZnS respectively. From Figure 10, a significant diffraction peak at 2θ = 28.7° which
corresponds to hexagonal wurtzite phase (002) of ZnS was observed. Another small peak appears
at 2θ = 30.8° which might be an intermediate phase between ZnO and ZnS. It can be termed as
ZnO1-xSx (zinc oxy-sulfide) or it could be possibly ZnS (101) phase. Figure 11 represents XRD
of two sulfidized ZnO films which were sulfidized for 4 hours for two different porous ZnO
parent samples. From Figure 11, I observed that after 4 hours of sulfidation ZnO has partially
converted to ZnS. Diffraction peaks at 2θ = 28.7° (ZnS 002) and the intermediate peak at 2θ =
30.8° matches with the previous result. From the plot, it is pronounced that the intensity of the
diffraction peak at 2θ = 28.7° is less for 4 hours sulfidation compared to 16 hours sulfidation.
This demonstrates that after 16 hours sulfidation most of the ZnO film has converted to
hexagonal ZnS. Figure 12, represents the XRD of the sulfidized ZnO peak that is 1 hour
sulfidized and Figure 13, represents the overlapping plot of XRD of the sulfidized samples for
ZnO Parent
19
(a)
(b)
Figure 10: a) XRD of sulfidized sample (16hr) b) cif file of ZnS
20
Two different sulfidized porous ZnO parent films which were sulfidized for 4 hours, better
crystallinity was achieved for the sample that was annealed in low oxygen atmosphere during the
synthesis of parent ZnO films.
(a)
(b)
Figure 11: XRD of sulfidized sample (4 hr.) (a) sample1 (b) sample2
21
From the XRD plot of 1 hour sulfidized film it can be concluded that ZnO has been partially
converted to ZnS which is consistent with the previous result.
Figure 13: Overlapping plot of parent and sulfidized samples for different time duration
Figure 12: XRD of sulfidized sample (1hr)
22
It is evident from all the XRD plot that time and temperature play crucial role in
conversion of ZnO to ZnS. The more the exposure time in sulfur environment, the greater the
intensity of ZnS (002) peak at any given temperature (180°C).The conversion process involves
in-diffusion of sulfur atom in appropriate ZnO crystal lattice site and out diffusion of oxygen.22
Also comparing the different XRD plots of parent and sulfidized samples it can be observed that
wurtzite ZnO (002) has converted to wurtzite ZnS (002) with preferred c-axis orientation. So, it
can be concluded that the conversion process of ZnO to ZnS does not alter crystal structure
although some paper reported sphalerite phase of ZnS along (111) diffraction peak after
conversion.19 However, for the intermediate peak at 2θ = 30.8°, two possible scenarios are there.
First, it can be another orientation of wurtzite ZnS phase ZnS (101) or it can be an in between
phase ZnO1-xSx. The value of x can be calculated using Vegard’s law. Figure 14 shows the lattice
parameter vs sulfur mole percent of the mixture which is helpful to determine the lattice
parameter at any given composition or if the lattice parameter of a mixed phase is known, the
chemical composition of the mixed phase can be predicted. The more the no of sulfur mole
percent converted, the higher should be the value of lattice parameter obtained because of larger
size of sulfur atom compared to oxygen atom. From the XRD plot of the sulfidized sample I
observed that position of the intermediate peak does not shift much depending upon the
sulfidation time. So, it is possible that a weak ZnS hexagonal phase developed in between which
disappears after the ZnO film undergoes complete sulfidation. The lattice parameters and
crystallite size of the sulfidized films are given in the Table 3 and Table 4 respectively. From the
Table 3, I see that lattice parameters of wurtzite ZnS matches the theoretical value c=
6.25730.23 The crystallite size of the sulfidized films can be calculated using Debye-Scherrer
equation as used earlier for the calculation of crystallite size of parent ZnO films. Apart from
23
temperature and time, the porosity of the parent film plays major role in sulfidation.
(a)
(b)
Research has shown porous ZnO sorbent with gaseous hydrogen sulfide accelerates sulfidation
process.15 The as-prepared ZnO parent films had inherent native defects. These defects affect the
migration of dopant atoms during diffusion. From Kirkendall’s work it is proved that diffusion
mechanism in substitution diffusion must propagate through vacancy formation.
Figure 14: Lattice parameter vs sulfur mole percent (a) 16 hours (b) 1 hour sulfidized films
24
Table 3: Lattice parameters of the sulfidized films
Table 4: Crystallite size of the sulfidized samples from porous ZnO
Studies has shown that for undoped ZnO, sulfidation can be achieved by two processes.
First, adsorption of H2S starts with formation of voids at ZnO/ZnS interface (i.e. the Kerkindall
effect).15 The sulfidation process accompanied by outward growth of ZnS where Zn2+ and O2- are
transferred to the ZnS/H2S gas surface. Secondly, sulfidation proceeds by inward growth
mechanism where S2- anions diffuse from outward surface to ZnO/ZnS interface internally. ZnO
deposition in oxygen atmosphere promotes more porous structure than the vacuum deposition.
Also, controlled growth of defects like oxygen vacancies and zinc interstitial affect sulfidation
process. By optimizing the background O2 pressure and post deposition annealing conditions,
oxygen vacancies can be controlled in the ZnO films. The sulfidation process discussed so far
were employed on more porous films with lot a of inbuilt oxygen vacancies. The
ZnS_sample 2() dhkl(Å) hkl a(Å) c(Å)
1_1 28.74 3.1048 0 0 2 3.81 6.21
3_1 28.59 3.1194 0 0 2 3.827 6.239
4_1 28.72 3.1058 0 0 2 3.81 6.212
ZnS from
k FWHM λ(Å) (rad) cos(rad) ß(rad) L(nm) h k L
ZnS 16hr 0.9 0.53545 1.5406 0.5013 0.9999 .00935 14.8 0 0 2
ZnS 1hr 0.9 0.51371 1.5406 0.4986 0.9999 .00897 15.4 0 0 2
25
photoluminescence effect of these porous ZnO films which shows strong defect emission peak
and weak UV emission peak supports the explanation. Analysis of Field emission scanning
microscopy and AFM (atomic force microscopy) revealed that increase of background oxygen
pressure converted porous crater -like structure into ZnO nanoparticle. Reduction of the pores
resulted from conversion of the pores into nanoparticles as oxygen pressure increased.24 Figure
15 represents the non-porous parent ZnO and sulfidized samples for 1hour, 16 hours and 32
hours respectively.
From figure 15, it was apparent that the parent ZnO film was highly crystalline in nature
with preferred c-axis orientation. The film was deposited in high oxygen pressure (4×10-1mbar)
and annealed in 600°c with background O2 pressure of 100 mbar which was 250 times higher
than the annealing pressure used for the previously deposited films. With the increase of
annealing temperature and pressure the crystallographic defects such as interstitial atoms, oxygen
vacancies, dislocation reduced in a significant rate.20 But this hinders the probability of
sulfidation as migration of the sulfur atoms happens through vacancy propagation. So, I observed
that after one hour of sulfidation, no new phase of ZnS appears, and that the ZnO (002) peak was
still there with significant reduction of intensity. After 16 hours of sulfidation, we can see that
two weak intensity peaks appear at 2θ = 28.6° and 2θ =30.8° which are ZnS (002) and the
intermediate peak ZnO1-x Sx/ ZnS (101). After 32 hours of sulfidation a strong peak of ZnS (002)
appeared at 2θ = 28.6° and a weak peak appears at 2θ =30.8° which corresponds to the
intermediate phase. If compared the sulfidation process of the previous ZnO films with this new
sample, it is noticeable that even after 16 hours of sulfidation not all the ZnO has converted to
ZnS. Additionally, 32 hours of sulfidation was also not long enough time for complete
conversion of ZnO to ZnS. Apart from that, two more weak intensity peaks appear at 2θ = 47.8°
26
and 2θ = 56.7° which corresponds to wurtzite ZnS (2 -1 0) and ZnS (2 -1 2) phases which means
the sulfidized film became polycrystalline in nature compared to preferential c-axis oriented
film.
The most plausible reason of slow rate of sulfidation must be vacancy formation which was not
prevalent in non-porous parent ZnO film annealed in high oxygen background. Indeed, the
photoluminescence spectra of this parent film has strong UV emission peak compared to the
Figure 15: XRD of a) nonporous parent ZnO, b) 16 hours sulfidized ZnO, c) 32 hours
sulfidized ZnO, and d) 1 hour sulfidized ZnO
27
previous ZnO films which implies that oxygen vacancies were reduced significantly that were
not helpful in migration of S2- ions through the lattice. The Figure 16 represents the comparison
of 16-hour sulfidation in 180°C for porous and non-porous films. Figure 17 represents the
overlapping Plot of sulfidation of non-porous ZnO films for different time duration.
Figure 16: comparison of sulfidation for porous and non-porous film for 16 hours
Figure 17: Overlapping XRD plots of sulfidized non-porous ZnO films for different time
duration
28
The lattice parameters of these non-porous parent sample and sulfidized samples can be
represented in Table 5 and Table 6 respectively. Table 7 shows crystallite size of the sulfidized
films from non-porous ZnO. I observed that the FWHM of the ZnS sample from non-porous
films are higher than the parent ZnO film. Crystallinity of sulfidized samples had degraded
compared to parent ZnO samples.
Table 5: Lattice parameter of non-porous parent ZnO sample
Table 6: Lattice parameter of ZnS sample 12 sulfidized for different amount of time.
Table 7: Crystallite size of non-porous sulfidized ZnO sample
ZnO non-porous Parent
Sample 2() d
12 34.45 2.601 0 0 2 3.192 5.203
ZnS_sample 12 2() dhkl(Å) hkl a(Å) c(Å)
12_3A 28.64 3.1144 0 0 2 3.821 6.229
12_3B 28.58 3.12 0 0 2 3.83 6.242
12_2B 34.48 2.6 0 0 2 3.19 5.198
Non-Porous parent sample ZnS
sample k FWHM λ(Å) (radian) cos(radian) ß(radian) L(nm) h k L
12_3A 0.9 0.4706 1.5406 0.2499 0.9999 0.008213519 16.8 0 0 2
12_3B 0.9 0.35945 1.5406 0.2494 0.999 0.006273586 22.1 0 0 2
12_2B 0.9 0.17756 1.5406 0.3008 0.9999 0.003099007 44.7 0 0 2
29
The average experimental lattice parameter c= 6.236 is matched with the reported value
c=6.25.23
SEM and EDS were performed on both the parent and sulfidized samples. Gold coating
was deposited on the parent sample and carbon coating was deposited on the sulfidized sample
for 30 seconds. Figure 18 shows SEM picture of the ZnO sample which was sulfidized for 16
hours. From the SEM picture it was apparent that the surface of parent ZnO sample was smooth
but after sulfidation of 16 hours the surface became rough and showed distribution like lamellar
structure. The figure 19 represents the EDS spectra of the sulfidized film.
Figure 18: SEM micrographs of (a) parent sample and 16hr sulfidized sample (b)40µm (c)
10µm (d) 20µm
30
From the figure below, the constitutional elements of the sulfidized film can be seen.
Raman spectroscopy was employed to study the vibrational energy modes in the system.
Raman spectroscopy can give information about the defects presents in the crystal structure,
bond force constant or the interaction between the atoms. The basic principle of Raman
spectroscopy is based on inelastic scattering of monochromatic light (laser source) after the
interaction with a sample that is Raman active. In inelastic scattering, the frequency of the
scattered light gets shifted up or down compared to the excitation frequency, which is called
Raman effect. A photon is absorbed by the material and jumps into higher vibrational energy
state. When it falls back to a lower energy level, three phenomena can happen. It can fall back to
same energy state as excitation energy which is called Rayleigh scattering. Or it can have a
transition to a lower or higher energy state, which are called Anti-Stoke or Stokes scattering
respectively. Stokes scattering was studied for both porous and non-porous parent ZnO and the
Figure 19: EDS spectra of the sulfidized sample
31
sulfidized samples.
Wurtzite ZnO with C6v point group symmetry has six Raman active modes.
Those are two A1, two E1, two E2 modes. Figure 20 shows Raman active modes for ZnO
nanoparticles. The frequencies of the fundamental modes of standard ZnO nanoparticles are
A1(TO) = 382 cm-1, A1(LO) = 574 cm-1, E2 (high) = 438 cm-1, E1(LO) = 583 cm-1. The second
order mode appears at 334 cm-1. Other modes that are reported in literature25 E2(low) = 101cm-1,
E1(TO) = 407 cm-1 which didn’t appear in the standard sample. The low frequency E2 mode
corresponds to the vibration of Zn atoms and the high frequency E2 mode is associated with
oxygen atoms.25 For strong preferentially oriented ZnO films when incident light is normal to the
surface, only A1(LO) and E2 modes are detected and other modes become forbidden.25
Figure 20: Raman active modes of ZnO nanoparticles
32
However, for the porous ZnO sample observed modes are 334, 378, 416, 574 cm-1. 334 cm-1 is
the second order vibrational mode as mentioned earlier. The peak at 378 cm-1 can be attributed to
A1(TO) symmetry mode which matched with the reported value in the paper by Wei et al.26 The
mode at 416 cm-1 corresponds to the sapphire substrate. The peak around 574 cm-1 can be
deconvoluted according to Lorentzian distribution. Figure 21 shows the overlapping plot of
Raman active modes of porous ZnO film, ZnO Nano powder and sapphire substrate. Figure 22
shows the peak fitting around 574 cm-1. The peak around ~ 575 cm-1 can be also attributed to
Raman active mode of sapphire substrate. Another peak observed around 581 cm-1 corresponds
to E1(LO) mode which according to Wei et al.26 comes from prevalence of oxygen vacancies in
the deposited film.
Figure 21: Overlapping plots of porous ZnO film with the ZnO Nano powder
33
The oxygen vacancy related defects can be resolved by annealing the films in oxygen rich
atmosphere. Wei et al.26 also reported another peak around 560 cm-1 which didn’t appear in
the porous ZnO film. They attributed this peak is observed due to incomplete oxidation of Zn in
their experiment. According to that paper ZnO contains large number of voids that can trap
interstitial zinc atoms which didn’t get oxidized. Figure 23 shows the Raman active modes of
non-porous ZnO thin film. Raman spectroscopy of the porous sulfidized films were studied as
well.
Wurtzite ZnS belong to space group C63mc having two formula units per primitive cell.
The Raman active modes for ZnS are one A1, one E1, two E2 and two B2 where A1 and E1 modes
are both Raman and IR active. 2E2 pair are only Raman active. Figure 24 shows the Raman
spectroscopy plot of a sulfidized film. The Raman spectrum of wurtzite ZnS was collected at room
temperature. A first order scattering at 347 cm-1 which corresponds to A1 and E1 symmetry LO
Figure 22: Deconvolution of peak of 574 cm-1 Figure 23: Raman active modes of non-porous
ZnO
34
modes was observed.
At ~274 cm-1 we observe two low frequency doublets assigned as A1/E1(TO) mode.
Xiong et al.27 reported A1/E1(TO) mode at 269 cm-1 for their ZnS nanowires which is shifted for
the thin film. They also reported another mode at 335 cm-1 which was not visible in the bulk
sample. That peak is due to surface phonon scattering. According to that paper in the ZnS
nanowire, surface optical phonon mode (SO) should not be observed because of perfect crystal
symmetry and smooth surface. This Raman active mode was observed in the sulfidized film. The
Figure 25 shows the surface optical phonon mode (SO) and E1(LO) modes. Figure 26 represents
Raman spectra of 1 hour sulfidized film. According to Brafman and Mitra E2 modes are reported
around 72 and 286 cm-1 28 which did not appear in the sulfidized film for this experiment.
Figure 24: Raman active mode of sulfidized sample for 16 hours
35
Deconvolution of the spectra around 340 cm-1 shows the surface active phonon mode ~ 335 cm-1.
Figure 25: Deconvolution of Raman active modes (LO/SO)
Figure 26: Raman spectra of ZnS sample (sulfidized for 1 hour)
36
The TO modes with A1 and E1 symmetry were reported at 273 cm-1 which matched with our
experiment. The LO modes of A1 and E1 symmetry were assigned at 351cm-1 which appears for
our experiment at 347 cm-1. Other group of researchers found TO mode around 274 cm-1 and LO
mode around 352 cm-1 at room temperature.29 Schneider and Kirby reported E1(LO) to be
around 348 cm-1. 30 Ebisuzaki and Nicol found A1/E1(LO) mode around 350 cm-1. 31 The ZnS
thin film Raman spectroscopy in our experiment is matched with reported values of other papers.
For the 2nd order Raman scattering the total region can be subdivided into three regions. i) low
frequency region (0-400 cm-1) where acoustic overtones are visible, ii) Intermediate frequency
region (400-540 cm-1) dominated by both optical and acoustic phonon, iii) high-frequency region
(540-750) dominated by optical overtone and combinations. In the low frequency region, we can
see a weak peak corresponds to acoustic overtone mode as LA (220 cm-1). The broad weak peak
around ~ 400 cm-1 can be assigned as TO+TA mode. The peak around ~619 and 671 can be
assigned as 2TO and 2LO modes. Cheng et al.32 reported LA overtone mode at 219 cm-1. And
the peaks around 622 and 676 cm-1 were assigned as 2TO and 2LO modes. Table 8 provides the
comparison between the experimental and reported value of the Raman active modes of porous
ZnO thin film. Table 9 represents the comparison of the reported and experimental values of the
Raman active modes of ZnS samples sulfidized from porous parent ZnO.
Table 8: Experimental and reported value of Raman active modes for porous ZnO thin film
Active Modes My Work Reported Modes( cm-1)
A1(TO) 378 380
A1(LO) 574 574
37
Table 9: Comparison of the Raman active modes of reported and experimental data of ZnS
In photoluminescence spectra of ZnO, there are two regions: Ultraviolet (UV) emission
band and visible emission band (green, yellow, blue, violet and orange). The UV emission is
characteristic emission of ZnO which happens due to exciton combination or near band edge
transition.33 The emission in the visible region is attributed to the intrinsic or the extrinsic defect
in ZnO.33 The Figures 27 and 28 shows the photoluminescence spectra of porous and non-porous
ZnO thin film. From the PL spectra of porous ZnO film, we can observe two emission peaks
around 380 nm and 510nm. The weak UV emission peak around 380 nm is attributed to exciton
recombination from conduction to valence band. The peak around 516 nm is related to defect
emission peak which is green emission. ZnO has several intrinsic defects that causes defect
emission peaks. i) oxygen vacancy (Vo),ii) zinc vacancy (Vzn), iii) interstitial oxygen (Oi), iv)
interstitial zinc (Zni), v) oxygen atom at position of zinc (Ozn), vi) zinc atom in position of
oxygen( Zno). 34 The formation energy of zinc vacancy (Vzn) is much larger than oxygen vacancy
(Vo) (5.4ev> 3ev). So, the probability of formation of oxygen vacancy is much higher than the
probability of zinc vacancy.
2TO 615 617 619
2LO 668 673 671
38
It is hypothesized that transition between singly ionized oxygen vacancies and photoexcited
holes is responsible for green emission.33
Figure 28: PL spectra of non-porous ZnO thin film
Figure 27: PL spectra of porous ZnO thin film
39
The yellow orange emission was attributed to the prevalence of interstitial oxygen and
dislocation present in the film.33 ZnO has intrinsic defects like oxygen vacancies, Zn interstitials
which decreases the probability of near band edge excitonic recombination compared to the non-
radiative recombination attributed to the different defects present in the crystal. If I compare this
PL spectra of the porous and the non-porous film, I can see that the UV emission peak ~380 nm
has improved significantly, and the defect related peak is not as intense as compared to the
porous film. This phenomenon can be attributed to the reduction of oxygen vacancies for the
non-porous film as it was grown and annealed in more oxygen rich atmosphere (100 mbar O2
pressure) than the porous film (annealed in 3x103 mbar O2 pressure). Zhaoyang et al.34 proposed
that defect green emission with higher intensity happens because of high oxygen pressure in the
sample. Ozn is more responsible for green emission compared to Vo. However, Jin et al.35
suggested that stoichiometry of ZnO thin films can be improved if the films can be grown in high
oxygen pressure. He suggested that ZnO thin films generally have lot of oxygen vacancies
resulting in poor stoichiometry. Annealing the films in high oxygen pressure has more improved
stoichiometry resulting in low oxygen vacancies and interstitial atoms. The XRD result of highly
crystalline quality and improved UV emission spectra of the non-porous ZnO film supports these
explanations. Wei et al.26 in the paper discussed that Zn vacancy acceptor level is around 2.6ev.
They observed blue emission from the PL spectra of their film which is of value 2.66 ev almost
close to 2.6ev. The source of blue emission is electronic transition from interstitial donor level
(Zni) to the acceptor level attributed to zinc vacancy (Vzn). Other group of researchers found that
concentration of oxygen vacancies (Vo) and zinc interstitial (Zni) is proportional to annealing
oxygen pressure as (PO2) -1/2 but the zinc vacancy concentration is proportional to (PO2)
1/2. Given
this fact, concentration of zinc vacancies can also increase the probability of the non-radiative
40
transition. Figures 29 and 30 represents the PL spectra of ZnO porous sample 3 and bulk ZnO
sample respectively. From the deconvoluted PL spectra of the porous ZnO sample, it can be
observed that the low intensity UV emission peak around 426 nm was red shifted in comparison
with the standard bulk sample.
Figure 29: PL spectra of porous ZnO sample 3
Figure 30: PL spectra of bulk ZnO sample
41
In deconvoluted PL spectra of the porous ZnO sample, the low intensity UV emission
peak around 426 nm which was red shifted with respect to the standard bulk sample, might be
possibly the blue emission26 which typically happens due to prevalence of oxygen vacancies in
the thin film sample. However, with the comparison of non-porous ZnO sample to the bulk
sample we find the UV emission spectra intensity is quite high as well as the defect spectra. The
most plausible explanation would be annealing in high oxygen pressure might have reduced the
oxygen vacancies and increased the zinc vacancies.35 So far, defects related emission of ZnO has
remained controversial for decades and the details of electronic transition are still not very clear
yet till date. PL of the sulfidized films also dependent on the sulfidation environment, time, and
the defects present in the crystals. The Figure 31 represents the PL spectra of the sample which
was sulfidized for 16 hours (Excitation wavelength 325 nm).
From the PL spectra of the sulfidized film I observed that the spectral shape was broad
and asymmetric. Researchers have reported blue emission band at 428 nm from ZnS. Becker and
Figure 31: PL spectra of the film sulfidized for 16 hours
42
Bard suggested the blue emission due to sulfur vacancies (S-2).36 Other group of researchers
suggested blue emission around 470 nm due to Zn+2 acceptor ions. Denzler et al. 37reported four
types of point defects which create intermediate trap levels inside the band gap. Zinc and sulfur
vacancies and interstitial atoms create four optical transition steps causing emission at longer
wavelength. The experimental value of optical band gap (2.8ev) of the sulfidized film was lower
than the theoretical optical band gap (3.8ev). Due to the increased surface to volume ratio
because of the lower crystallite size the probability of excitonic transition decreases compared to
the non-radiative surface transition due to induced trap levels. Figure 32 represents the plot of
another porous sample which was sulfidized for 1 hour. From this plot, the appearance of blue
emission band around 428nm .
A weak shoulder peak around 523 nm was seen which can be attributed to the recombination of
the electron hole pair on the surface traps that lies between the band gap. Hu et al.38 observed
Figure 32: PL spectra of the porous sulfidized film (1 hour)
43
emission in green region and attributed the self-activated centers containing vacancies and the
interstitial states as the surface traps. Figure 33 ,34 and 35 represents the PL spectra of the non-
porous films which were sulfidized for 16 hours,32 hours and 1 hour respectively.
Figure 34: PL spectra of ZnS sample12(32 hour sulfidized)
Figure 33: PL spectra of ZnS sample12(16 hour sulfidized)
44
From the Figure of non-porous film, which was sulfidized for 16 hours, blue emission
band attributed to sulfur vacancy (S-2) appeared at 450 nm instead of 428 nm as resulted for the
porous sulfidized films. The reason for the red shift in this emission band has not yet been
reported. Though Bhattacharjee and Lu reported red shift of the blue emission band due to
increase of annealing temperature.39 The reason for red shift in our samples needs more
investigation. For the film which was sulfidized 32 hours no blue emission band around 428nm
was observed. Most plausible explanation would be less sulfur vacancy due to more exposure
time in sulfur rich environment. Two other bands around 550 nm and 650 nm appear for surface
traps present in the band gap.
From the deconvoluted spectra of the film which was sulfidized for 1 hour, presence of
characteristic UV emission peak of ZnO around 390 nm was evident and the defect emission
peaks around ~ 540 nm and 640 nm which could be caused by prevalence of surface defect states
Figure 35: PL spectra of sample12 (1 hour sulfidized) (a) UV emission and (b) defect emission
45
inside the band gap were also observed.
From the XRD, Raman spectra, and PL spectra analysis it is observed that ZnO thin films
are converted to ZnS thin films which is highly dependent upon sulfidation time and
temperature. Sulfidation time and temperature play an important role controlling the
stoichiometry of the films. At any certain temperature the sulfidation is dependent on the
exposure time on the sulfur environment. The more the exposure time, the more layers of ZnO
will get sulfidized. However, vacancy plays an important role for migration of the sulfur atoms
through the crystal lattice. Whether the parent ZnO film is porous or not, PL and Raman
spectroscopy can provide lot of details about the stoichiometry of the parent ZnO films. For
instance, the strong UV emission intensity defines less oxygen vacancy and better stoichiometry
of the ZnO films. Though sulfidation may not be very fast because lesser vacancies make it
difficult for the sulfur atoms to move through the lattice and replace the oxygen. For non-porous
parent films, it needs more exposure time in the sulfur environment to sulfidize the films when
porous defective ZnO films might have poor stoichiometry but those got better sulfidized for
various reasons discussed through each section. Additionally, as diffusion of S2- anions are
essentially temperature dependent it would be worthwhile to study the onset of diffusion process
by varying the temperature of sulfidation while keeping the time of sulfidation constant. As part
of my experiment, I have explored the temperature dependence on the sulfidation process which
caused structural variation of the sulfidized films through XRD analysis. Figure 36 represents the
XRD of porous ZnO parent sample and the comparison of XRD of the sulfidized films which
were sulfidized for 4 hours at 100°C and 180°C.
From the figure I saw that the sulfidized film at 100°C has an evident peak of wurtzite
ZnO at around 2θ ~ 34.5° whereas the film which was sulfidized at 180°C had no presence of
46
ZnO. A new phase of wurtzite ZnS (002) developed around 2θ ~ 28.6° followed by complete
sulfidation of porous ZnO at 180°C. No intermediate peak around 2θ ~ 30.7° which corresponds
to ZnO1-x Sx / ZnS (101) was observed. A weak peak of ZnO1-xSx/ ZnS (101) grew around 2θ ~
30.7°was also seen.
(a)
(b)
Figure 36: (a)XRD of a porous parent ZnO (b) Overlapping plot of 4 hours sulfidation at
180°C
47
However, the intensity of the peak corresponding to wurtzite ZnO (002) has significantly
got reduced. Hence it can be deduced that activation energy of diffusion of S2- ions is strongly
dependent on temperature. More in depth analysis could be done on temperature dependence of
activation energy of diffusion of anions and sulfidation kinetics. Additionally, the structural and
optical properties can also be altered by varying the temperature of sulfidation.
Conclusions
ZnO-ZnS thin film heterostructures have been successfully synthesized by pulsed laser
deposition and hydrothermal method where oxygen in the PLD grown ZnO thin films got
replaced with sulfur by diffusion mechanism. XRD result shows partial and complete conversion
from ZnO to ZnS thin film depending on the time of the sulfidation. Porosity and oxygen
vacancy play important role in sulfidation process. In general, the more vacancy rich and porous
the parent ZnO film, the more oxygen gets replaced by sulfur because sulfur atoms can easily
migrate through vacancies. The crystallinity of the sulfidized films turn out poor compared to the
parent films. Surface irregularities and uneven distribution of particles were clearly visible in the
SEM pictures. The PL spectra of the ZnO films give information about the stoichiometry of the
films and the defects that are present in the parent films. The optical band gap of the parent ZnO
films turned out to be ~ 3.18ev which is less than the theoretical reported value 3.36ev. Blue
emission band was observed in the porous ZnO films that were sulfidized for different amount of
time. Sulfur vacancies are the plausible reason for blue emission of the sulfidized films as
reported in some of the papers. From the Raman spectroscopy of the parent and the sulfidized
films we could see the Raman active modes for wurtzite ZnO and ZnS structure. More analysis
needs to be done through theoretical calculation and simulation to understand the temperature
48
and time dependency of diffusion mechanism which enables migration of sulfur atoms through
the crystal lattice for replacement of oxygen with sulfur in hydrothermal synthesis.
49
NANOPARTICLES HETEROSTRUCTURES
Abstract
Small ZnO nanoparticle is emerging as industry standard solid sorbent for removal of
sulfur-based compound from fossil fuel. While sulfidation is limited to the surface of bulk ZnO,
small ZnO nanoparticle is more effective for absorbing sulfur more efficiently. The structural
and optical characteristics of sulfidized ZnO nanoparticle are studied in this paper. The wurtzite
ZnO undergoes complete sulfidation to a new crystallographic structure comprised of mixed
phase of sphalerite and wurtzite ZnS. The exchange of anion happens through diffusion process
and contribute to substitutional doping. The sulfidized product nanoparticle had different size
than the crystallite size of parent sample. The Photoluminescence spectra revealed more defect
states within the band gap which is evident from the defect emission spectra in the visible region.
Introduction
ZnO nanoparticles have a wide range of applications such as energy storage, catalysis,
coating and pigments, environmental remediation and many others.40 ZnO has been used as a
solid sorbent for removal of polluting sulfur from industrial petroleum-based fuel. Very few
papers previously reported that sulfidation of ZnO bulk powder is limited to the surface whereas
for the nano powder the entire ZnO can undergo complete sulfidation because of high sorbent
capacity. ZnO nano powder can act as sorbent to scavenge sulfur compound such as hydrogen
sulfide from petroleum-based products and fuels.41 The reaction can be represented as
50
ZnO+H2S→ ZnS+H2O. The oxygen ions (O2-) in the lattice get replaced by sulfur ions(S2-).
Specifically, for the nanoparticles, the more exposed surface area enables enhanced kinetics for
sulfidation reaction at relatively low temperature. Banerjee and Jain reported much lower
desulfurization temperature compared to conventional operating temperature in the range 650-
800°C.40 The mechanism of sulfidation process of nanoparticles has critical difference compared
to the bulk material, even the few nm differences in size may have effect in sorbent design. In a
study, Park et al.42 reported sulfidation of hexagonal ZnO nano-particles of size 14 nm at 235°C
using hexamethyldisilathiane. The reaction involved exchange of O2- with S2- in the crystal
lattice. They also reported faster outward diffusion of Zn2+ compared to inward diffusion of S2-
which is known as nanoscale Kirkendall phenomenon which resulted hollow nanocrystalline
structure. Another group of researchers reported growth of a protecting layer of ZnS on the
surface of ZnO nanoparticle at relatively lower temperature. Where first stage of sulfidation
reaction is relatively easy and rapid at low temperature, the bulk reaction is tough because of
diffusion resistance.15 Also, ZnO based sorbent can absorb more H2S if large number of atoms
are exposed on the surface. To increase sulfur intake by ZnO based sorbent at low temperature,
complete sulfidation of bulk ZnO is required. For this complete sulfidation, diffusion through
primary ZnS layer must be accelerated. However, very less detailed information is available
about the sulfidation kinetics. Till this date it has been established that inward growth leads to
formation of ZnS layer where sulfur diffuses from external ZnS/H2S layer to the ZnS/ZnO
interface and oxygen atoms which are transferred to inverse direction can be combined to the
proton and resulted in water molecules.15 However, new data has also emerged which has shown
sulfidation of ZnO nanorods with H2S in liquid medium. It has shown Kirkendall effect which
resulted in outward growth of ZnS layer. In this process, diffusion of Zn2+ and O2- happens
51
through ZnS layer to the external ZnS/H2S interface. Zn2+ cation combine with sulfur and O2-
combines with protons to result in H2O molecules. The figure 37 represents schematic of ZnS
growth during sulfidation of ZnO.
In my experiment, I have studied the sulfidation of ZnO nanoparticles for 16 hours ,6
hours and 1 hour using X-ray crystallography and optical spectroscopy. Structural and optical
characteristics were analyzed before and after sulfidation.
Experimental Methods
ZnO-ZnS CSNPs have been synthesized using hydrothermal synthesis. 200mg thiourea
mixed with 40ml DI water served as precursor solution. 60 mg of ZnO nano powder was well
dispersed in 0.066 M thiourea solution by means of probe sonicator and sulfidized for 16 hours,
6 hours and 1 hour, respectively at 180°C in an autoclave. After sulfidation, the autoclave was
cooled down to the room temperature and sulfidized samples were recovered by centrifugation.
The phase and crystallinity of the sulfidized samples were characterized by x- ray diffractometer
(Bruker AXS D8) equipped with high-resolution detector using Cu-Kα source with wavelength
Figure 37: Mechanism of ZnS growth during ZnO sulfidation of inward and outward growth
52
of 1.5406. The operating voltage and current were 40 KV and 40 mA respectively. θ-2θ scan
was performed from 20° to 80° after optimization of and Z axis scan. The data were analyzed
using Origin pro 8.5. Surface morphology and elemental analysis of the sulfidized
heterostructures were done using scanning electron microscopy (Fei Quanta 2000) and energy
dispersive x-ray spectroscopy (EDS) that has been equipped with in the scanning electron
microscope.
Raman spectroscopy was conducted on both as grown and sulfidized nanoparticle to
study the vibrational modes present in the system. The As-grown ZnO and sulfidized samples
were characterized using 532nm green laser by Horiba Labram HR Raman PL system.
PL spectroscopy was also conducted where the samples were irradiated with 325 nm, He-
Cd laser source at room temperature. Excitation energy was above the band gap of the material.
The samples were flattened on a metal bar of tungsten carbide which contributes minimum
background signal. The 15x NUV-lens was used for focusing on the localized spot and
measurements were taken at multiple spots. Beam splitter UV splits the scattered light from the
samples according to the wavelength. The scan was taken in the range of 350-750 nm while
acquisition time was 20 seconds, and 20 scans were taken in average in that wavelength range.
Both Raman and PL spectroscopy data were analyzed using Origin pro 8.5 software.
Results and Discussion
Figure 38 represents the XRD pattern of the wurtzite ZnO nano powder which reveals
that the sample was polycrystalline in nature. Figure 39 represents the cif file of wurtzite ZnO
which matches with the experimental values. From the XRD data of the ZnO wurtzite sample I
53
can conclude that the sample was polycrystalline in nature which was not preferentially oriented
as opposed to the thin films. Lattice parameters of the ZnO nanoparticles found out to be
consistent with the theoretical value. Table 10 represents the lattice parameters of wurtzite ZnO
which are consistent with the theoretical value.
Table 10: Lattice parameters of wurtzite ZnO
2θ (°) dhkl h k l a() c()
31.67 2.8229 1 0 0 3.259 5.321
34.3 2.612 0 0 2 3.199 5.224
36.11 2.485 1 0 1 3.244 5.298
47.358 1.918 1 0 2 3.227 5.271
56.45 1.628 2 -1 0 3.25 5.307
Figure 38: XRD of wurtzite ZnO nano powder
54
The values of the average lattice parameters were calculated as a = 3.23 and c = 5.28.
Figure 40 shows XRD plot of the ZnS nano powder which is sulfidized for 16 hours. From
the XRD plot of the ZnS nanoparticles sulfidized by hydrothermal synthesis for 16 hours, wurtzite
ZnO nanoparticles have been converted to ZnS nanoparticles comprising of mixed phase of
wurtzite and sphalerite structures. This is supported by the complete loss of diffraction peaks of
wurtzite ZnO and appearance of diffraction peaks corresponding to (002) plane of wurtzite ZnS
and (200) plane of sphalerite ZnS and other peaks of mixed phases as mentioned in the figure.
Although researchers Ma et al.41 reported formation of sphalerite phase than formation of the
mixed phase as in case of our experiment.
Figure 39: Cif data of wurtzite ZnS
55
The inter planar spacing corresponding to 2θ ~ 28.5° turned out to be 3.129 which is very
close to the reported value for zinc-blende phase. Although, some researchers reported formation
of the mixed phase of wurtzite and sphalerite phases.15 The particle size of synthesized ZnS
nanoparticles can be calculated using Debye-Scherrer equation. The XRD plots of the 6 hours
and 1-hour sulfidation are represented below in Figures 41 and 42, respectively. These plots are
consistent with the plot of 16 hours sulfidized ZnO NPs. From the XRD plots of sulfidized NPs it
was observed that the positions of the diffraction peaks of the sulfidized nanoparticles for
different times were very little different although the intensities were comparatively low for 6
hours and 1 hour sulfidized nanoparticles than the 16 hours sulfidized one. Also, omission of the
wurtzite (100) phase and sphalerite (200) phase in 1 hour sulfidized nanoparticle were evident.
Further investigation is needed in order to study the sulfidation kinetics for varying amount of
Figure 40: XRD of ZnS nano powder sulfidized for 16 hours
56
Figure 41: XRD of ZnS nano powder (6 hours sulfidized)
Figure 42: XRD of ZnS nano powder (1 hour sulfidized)
57
Figure 43 shows the overlapping plot of the sulfidized nanoparticles for different hours.
(a)
(b)
Figure 43: a) Overlapping XRD plot of ZnO nano powders sulfidized
for different hours b) comparison with ZnS cif data
58
SEM and EDS were performed on the sulfidized nanoparticles. Figure 44 represents SEM
micrograph of pure ZnO and sulfidized NPs of 16 hours and 6 hours, respectively. From the
SEM micrograph of pure ZnO NP, clustered and cloudy particles were observed. The 1 hour
sulfidized picture is also quite clustered. Distinct particles are almost invisible.
From the SEM picture of ZnO nanoparticle, I observed cloudy and blurry crystals of
ZnO. Researchers have reported this kind of micrograph due to presence of hydroxyl group of
ions present in aqueous solution of ZnO. This holds true for the sulfidized NPs of 1 hour. This
must have been very small amount of time for the nanoparticles to be free of the hydroxyl ions.
Figure 44: SEM micrographs of (a) pure ZnO nanoparticles (b) ZnS nanoparticles (1 hour
sulfidized), (c) 6 hours, and (d) 16 hours
59
Additionally, smaller particle size for 16 hours sulfidized nanoparticles compared to the 6 hours
sulfidized one can be seen. Some researchers reported the smaller the particle size the more
sulfidation will take place. Further, formation of cavities or voids within the structure were also
observed. This might have happened due to outward growth of ZnS followed by the diffusion of
Zn2+ and O2- ions to the external interface of ZnS/H2S(liquid) as known as Kirkendall
effect.15Also, the 6 hours sulfidized nanoparticle seems to be appeared more clustered. Rouhi et
al.43 reported in their paper that voids have been formed at the ZnO- ZnS interface for partially
sulfidized ZnO. According to the paper the paper the voids formed due to zinc and oxygen
vacancies in ZnO which coalesce together and aid the formation of voids at the interface of ZnO
and ZnS. As the interface between ZnO and ZnS phase is reduced it limits the diffusion of zinc
and oxygen into ZnS phase. Diffusion of zinc and oxygen atoms are very important for
sulfidation reaction which has direct impact on sulfidation kinetics.43 The EDS spectra gives the
information about the compositions of the constituent elements. Figure 45 shows the
constitutional analysis of the 16 hours and 6 hours sulfidized nanoparticles respectively.
(a) (b)
Figure 45: EDS analysis of sulfidized NPs for (a) 16 hours and (b) 6 hours
60
The Raman spectra of the sulfidized CSNPs have been studied. Figure 46 shows the
Raman active modes of 1 hour ,6 hours and 16 hours sulfidized ZnO NPS.
Figure 46: Raman active modes of sulfidized NPs for (a)16 hours, (b) 6 hours, and (c) 1 hour
61
From these spectra fewer number of Raman active modes compared to the sulfidized thin films
were observed. A first order scattering around 344 cm-1 related to A1 and E1 symmetry mode
was observed for all sulfidized NPs. However, for the sulfidized thin film samples this mode
appeared around 347 cm-1. Brafman and Mitra reported this mode at 351 cm-1 while other group
of researchers observed this mode around 352 cm-1 at room temperature.28,29 A second order
Raman active mode was observed 619 cm-1 for 16 hours and 6 hours sulfidized NPs. Cheng et
al.32 reported 2nd order TO mode around 622 cm-1. However, this mode was not observed for 1
hour sulfidized NPs. Therefore, it can be concluded that after significant amount of time during
the sulfidation 2nd order TO mode can be observed. More research needs to be conducted to
understand the omission of the Raman active modes of sulfidized NPs compared to the sulfidized
thin films.
To investigate the optical properties of the ZnO-ZnS core shell nanoparticles, PL
spectroscopy was performed at room temperature. Figures 47 and 48 represent the PL spectra of
the ZnO-ZnS CSNPs sulfidized for 16 hours and ZnO nanoparticles for reference. The samples
were excited using 325 nm He-Cd laser. Several researchers have reported blue emission of ZnS
nanoparticles under UV excitation at 428 nm. Becker and Bard attributed this blue emission due
to the sulfur (S2-) vacancies.36 Murase et al.44 suggested blue emission band around 470 nm to
Zn2+ acceptor ions. Yanagida et al.45 have reported defect related emission at 420 nm. Although
in our experiment no peak is visible around 428nm related to excitonic emission as opposed to
ZnO-ZnS thin films. In ZnO-ZnS NP heterostructure, PL spectra appear at significant lower
energy level than the theoretical optical band gap energy (3.54ev for ZB and 3.91ev for Wurtzite
ZnS). This supports favorable transition from energy states inside the band gap. Chen et al.46
attributed PL peaks at the lower energy level than the band gap to the surface states. According
62
to the author, as the particle size decreases, the surface to volume ratio and surface states
increase. It was also proved by their thermo-luminescence measurement.
Figure 47: PL spectra of 16 hours sulfidized nanoparticles
Figure 48: PL spectra of ZnO NP
63
As a result, the probability of excitonic recombination decreases via non-radiative surface states
recombination. So, the defect emission peaks around~ 525, 583 and 625 nm might have resulted
from prevalence of surface defect states present in the ZnO sample that is sulfidized for 16 hours.
Figure 49 shows defect states of ZnS NP. Figures 50 and 51 represent plots of PL of ZnO sample
that are sulfidized for 6 hours and 1 hour.
In the literature PL peaks ~ 2.90ev/2.91ev is assigned to singly ionized sulfur vacancies.47 The
emissions ~ 2.76/2.77ev are ascribed to Zn vacancies. Further researchers reported spectra ~
2.31ev due to emission from the sulfur vacancies to Zn vacancies.47 The emission spectra of 6
hours and 1 hour sulfidized ZnO NPs are consistent with the result of 16 hours sulfidation. These
plots also suggested PL emission peaks around ~ 2.36ev, 2.13ev and 2ev. So, in my experiment
the emission spectra ~ 2.36ev might have resulted from transition to Zn vacancies from sulfur
vacancies.
64
Also as suggested from the XRD data of the ZnO-ZnS NP heterostructure, majority of the
ZnO was converted to ZnS preferably with sphalerite structure, more time and temperature-
controlled synthesis needs to be done in order to understand sulfidation kinetics and defect
emission of the heterostructure.
65
Conclusions
ZnO-ZnS core-shell nanoparticles had been successfully formed due to sulfidation of
ZnO nanoparticles at 180°C for varying amount of time. At 180°C voids might have formed at
ZnO-ZnS interface which facilitated ZnS nanoparticle formation. Previous researchers have
reported this kind of growth during sulfidation of ZnO nanorods.48 This is evidence of Kirkendall
effect which suggests outward growth of ZnS layer. Several studies have shown that while
reaction has been carried out between ZnO and H2S at 350°C, voids still formed after partial
transformation, but are absent for completely transformed ZnS. Other researchers reported
sulfidation of ZnO nanoparticles where O2- ions are substituted by S2- ions.48 Mixed phase of
ZnS wurtzite and zinc-blende formed during the sulfidation process. The more S2- are available
more ZnS domains grow combining with Zn2+ ions.48 From the XRD analysis we confirm that
initial wurtzite structure of ZnO is not retained after nucleation-growth process of forming ZnS.
It has also been found that the difference between the unit cell of ZnO and ZnS induces lattice
strain. The PL spectra confirm prevalence of defect states inside the band gap which gives rise to
emission in the visible region.
66
CONCLUSIONS
Sulfur doped heterostructures of ZnO-ZnS have been successfully synthesized at 180°C
using PLD and hydrothermal synthesis where oxygen is exchanged with sulfur for both the thin
film and nanoparticles. Structural characterization has been performed using XRD and SEM
which indicate the partial to complete exchange of oxygen ion with sulfur ion in ZnO through
hydrothermal method. It also shows that oxygen vacancies play an important role in the
sulfidation process. ZnS might have grown through void formation at the ZnO-ZnS interface. For
the thin film we have seen that initial wurtzite crystal structure of the thin film has retained and
an intermediate phase known as zinc-oxysulfide might have grown after sulfidation where for the
nanoparticles, a mixed phase of sphalerite and wurtzite ZnS has formed after sulfidation. From
the Raman spectroscopy one can observe the Raman active modes of ZnO and ZnS, though
further investigation is needed for determination of the Raman active modes of ZnO1-xSx. PL
spectra of the pure and sulfidized thin films and the nanoparticles has provided information about
the defect states which causes electronic transition in the visible region. PL spectra also helped
us find the band gap of the ZnO-ZnS heterostructures. Researchers have observed
ferromagnetism of non-stoichiometric ZnS microspheres. Therefore, along with optoelectronic
applications, the ZnO-ZnS thin film/ core-shell nano- heterostructure can be exploited to study
novel ferromagnetic behavior in future.
67
REFERENCES
1 G. Bastard, Wave mechanics applied to semiconductor heterostructures, 1990.
2 M. Jaros, Wave Mechanics Applied to Semiconductor Heterostructures, Taylor & Francis,
1991.
3 M.-Y. Lu, J. Song, M.-P. Lu, C.-Y. Lee, L.-J. Chen and Z. L. Wang, ZnO−ZnS
Heterojunction and ZnS Nanowire Arrays for Electricity Generation, ACS Nano, 2009,
3(2), 357–362.
4 R. Dingle, H. L. Störmer, A. C. Gossard and W. Wiegmann, Electron mobilities in
modulationdoped semiconductor heterojunction superlattices, Appl. Phys. Lett., 1978,
33(7), 665–667.
5 L. J. Lauhon, M. S. Gudiksen and C. M. Lieber, Semiconductor nanowire
heterostructures, Philos. Trans. R. Soc. London. Ser. A Math. Phys. Eng. Sci., 2004,
362(1819), 1247–1260.
6 S. R. White and L. J. Sham, Electronic Properties of Flat-Band Semiconductor
Heterostructures, Phys. Rev. Lett., 1981, 47(12), 879–882.
7 V. I. Tolstikhin, Optical properties of semiconductor heterostructures